The Purge old results service removes outdated matching metadata generated by the add-on. This helps improve matching performance and reduces storage usage.
Note that matching metadata is stored in a technical dataset created by the add-on. A separate dataset is created for each table where matching is executed. The following naming convention is used for datasets: <datasetName>_<tableName>_MDS.
When you run the Purge old results service:
Redundant suspect relationships are removed between records that were flagged as false-positive. This impacts the MatchingResult and FalsePositiveDecision metadata tables.
Outdated matching results where the source and target record appear in multiple matching processes are cleaned. This impacts the MatchingSelection and MatchingResult metadata tables.
The MatchingSelection table is truncated.
Records that are orphaned for any of the following reasons are removed from the RecordMetadata metadata table:
If record was, but is no longer linked to a golden record (for example if the golden was deleted).
If a record was deleted in the main table, but still exists in the RecordMetadata table.
If this service is run on a Business Object, all related child metadata tables are also purged.